AUTHOR=Mateo Carlos M. , Corrales Juan A. , Mezouar Youcef 

TITLE=Hierarchical, Dense and Dynamic 3D Reconstruction Based on VDB Data Structure for Robotic Manipulation Tasks

JOURNAL=Frontiers in Robotics and AI

VOLUME=Volume 7 - 2020

YEAR=2021

URL=https://www.frontiersin.org/journals/robotics-and-ai/articles/10.3389/frobt.2020.600387

DOI=10.3389/frobt.2020.600387

ISSN=2296-9144

ABSTRACT=This paper presents a a reviewed approach to implement hierarchical, dense and dynamic reconstruction method based on VDB (variational dynamics B+ Trees) data structure for robot tasks.
Scene reconstruction is done by the integration of depth-images using the Truncated Signed Distance Field (TSDF). 
Nowadays, dense reconstruction domain is ruled by three major space representations, complete volumes, hashing voxels and hierarchical volumes. 
Here, we propose design the reconstruction method based on dynamic trees can provide similar reconstruction result than current stat-of-art methods, but with a direct multi-level representation at expenses of just a slightly higher computational cost, being still real-time. 
Additionally, this representation provide two major advantages against the other, hierarchical and unbounded space representation.
The proposed method is optimally implemented to be used on a GPU architecture, exploiting the parallelism skills of this hardware.
A series of experiments will be presented to prove the performance, qualitatively, of this approach in a robot arm platform.